Diagnostic tests can guide therapeutic choices in chest pain
In patients with stable chest pain, the burden of low-attenuation noncalcified plaque on coronary CT angiography is a better predictor of future myocardial infarction risk than a cardiovascular risk score, an Agatson coronary artery calcium score, or angiographic severity of coronary stenoses, Michelle C. Williams, MBChB, PhD, reported at the joint scientific sessions of the American College of Cardiology and the World Heart Federation. The meeting was conducted online after its cancellation because of the COVID-19 pandemic.
Dr. Michelle C. Williams
These findings from a post hoc analysis of the large multicenter SCOT-HEART trial challenge current concepts regarding the supposed superiority of the classic tools for MI risk prediction, noted Dr. Williams, a senior clinical research fellow at the University of Edinburgh.
Indeed, it’s likely that the current established predictors of risk – that is, coronary artery calcium, severity of stenosis, and cardiovascular risk score – are associated with clinical events only indirectly through their correlation with low-attenuated calcified plaque burden, which is the real driver of future MI, she continued.
Histologically, low-attenuated noncalcified plaque on coronary CT angiography (CCTA) is defined by a thin fibrous cap, a large, inflamed, lipid-rich necrotic core, and microcalcification. Previously, Dr. Williams and her coinvestigators demonstrated that visual identification of this unstable plaque subtype is of benefit in predicting future risk of MI (J Am Coll Cardiol. 2019 Jan 29;73[3]:291-301).
But visual identification of plaque subtypes is a crude and laborious process. In her current study, she and her coworkers have taken things a giant step further, using commercially available CCTA software to semiautomatically quantify the burden of this highest-risk plaque subtype as well as all the other subtypes.
This post hoc analysis of the previously reported main SCOT-HEART trial (N Engl J Med. 2018 Sep 6;379[10]:924-933) included 1,769 patients with stable chest pain randomized to standard care with or without CCTA guidance and followed for a median of 4.7 years, during which 41 patients had a fatal or nonfatal MI. At enrollment, 37% of participants had normal coronary arteries, 38% had nonobstructive coronary artery disease (CAD), and the remainder had obstructive CAD.
In a multivariate analysis, low-attenuation noncalcified plaque burden was the strongest predictor of future MI, with an adjusted hazard ratio of 1.6 per doubling. This metric was strongly correlated with coronary artery calcium score, underscoring the limited value of doing noncontrast CT in order to determine a coronary artery calcium score when CCTA is performed.
Low-attenuation plaque burden correlated very strongly with angiographic severity of stenosis, and only weakly with cardiovascular risk score, perhaps explaining the poor prognostic performance of cardiovascular risk scores in SCOT-HEART and other studies, according to Dr. Williams.
Patients with a low-attenuation noncalcified plaque burden greater than 4% in their coronary tree were 4.7 times more likely to have a subsequent MI than were those with a lesser burden. The predictive power was even greater in patients with nonobstructive CAD, where a low-attenuation noncalcified plaque burden in excess of 4% conferred a 6.6-fold greater likelihood of fatal or nonfatal MI, she observed.
Two things need to happen before measurement of low-attenuation noncalcified plaque via CCTA to predict MI risk is ready to be adopted in routine clinical practice, according to Dr. Williams. These SCOT-HEART results need to be validated in other cohorts, a process now underway in the SCOT-HEART 2 trial and other studies. Also, improved software incorporating machine learning is needed in order to speed up the semiautomated analysis of plaque subtypes, which now takes 20-30 minutes.
Dr. Williams reported having no financial conflicts regarding her study, funded by the National Health Service.
In conjunction with her virtual presentation at ACC 2020, the SCOT-HEART study results were published online (Circulation. 2020 Mar 16. doi: 10.1161/CIRCULATIONAHA.119.044720. [Epub ahead of print]).
SOURCE: Williams MC et al. ACC 2020, Abstract 909-06.